SpeechPro Releases Speaker Recognition SDK for ID Applications

SpeechPro, the U.S. subsidiary of Speech Technology Center, yesterday released its VoiceGrid software development kit (SDK) for the U.S. market.

The VoiceGrid SDK provides the same tools and algorithms available within the VoiceGrid line of turn-key biometric voice identification applications but with the ability for business partners with development resources to integrate the technology into their own platforms.

The VoiceGrid SDK provides a unique set of capabilities for the voice biometric platform. The average time of biometric feature extraction is 10 seconds, the system performs more than 3,000 comparisons of voice models per second on a single core and can support a database of 100,000 voice prints within only 2 gigabytes of RAM space. It has client-server architecture allowing access over HTTP or with the client library implemented for most commonly used programming languages. The server-side of SDK can run on Windows and Linux, on both 32 and 64-bit architectures.

This capability allows a wide range of developers to create the solutions of any scale, from desktop applications to large scale distributed identification systems, with almost real-time response times providing fast and accurate results of biometric searches to the end users.

Alexey Khitrov, president of SpeechPro USA, stated, "The VoiceGrid SDK represents an important step in assisting us to grow the U.S. market through strategic business partnerships by placing our technology in the hands of partners who already have relationships with customers within the vertical markets in which voice identification is well suited to add value and increase effectiveness to investigative, security and public safety efforts."

"We quickly identified that within the U.S. market, allowing business partners to incorporate voice identification technology into their existing applications in many situations, rather than trying to promote separate and potentially disparate turn-key systems was the right move," said William Michael, national sales director of SpeechPro USA. "We have already delivered several SDKs to well respected business partners who operate within the U.S. law enforcement market, and we only expect this type of demand to grow as word gets out about the SDK."

The VoiceGrid SDK allows to partners to develop systems for both voice databases and real-time identification systems. The price of the SDK can be offset 100 percent against licensing costs once the partner has developed a deployable application. Licensing pricing is based on the database size and the number of cores used in the final application.